Among the numerous functions of thyroid hormones is their integral role in glucose homeostasis. Hypothyroidism has been associated with poor glycemic control in diabetic dogs but has been suggested to be an uncommon cause of insulin resistance.1 However, diabetes mellitus was the most common concurrent disease in a retrospective study2 of dogs with hypothyroidism, and hypothyroidism was one of the most commonly diagnosed concurrent disorders in dogs with diabetes mellitus.3 Findings in dogs with experimentally induced hypothyroidism suggest the presence of insulin resistance, as indicated by markedly exaggerated plasma insulin concentrations after PO or IV glucose tolerance tests.4–7 Low SI and impaired glucose use associated with hypothyroidism have been identified in other animal species as well.8,9
Whether the glucose intolerance in hypothyroidism is the direct result of thyroid hormone deficiency or whether hormones with counter-regulatory effects on insulin mediate the development of insulin resistance is unclear. For example, in hypothyroid insulin-resistant women, plasma concentrations of glucagon, epinephrine, cortisol, and GH were high before hormone replacement therapy, indicating that these hormones may contribute to the development of insulin resistance.10 Growth hormone excess has been linked to insulin resistance in dogs, cats, and people,11–13 and recent studies14–16 have revealed high GH concentrations in hypothyroid dogs. Obesity, which is a common consequence of the reduced metabolism of hypothyroid dogs, may also negatively affect SI and glucose tolerance in dogs.17
Although standard glucose tolerance tests provide a general assessment of glucose tolerance and insulin secretion, more dynamic evaluation of insulin sensitivity and glucose disposal is provided by other experimental techniques. Results of the FSIGT with minimal model analysis correlate well with those of the reference (gold) standard, the euglycemic hyperinsulinemic clamp.18 The model not only calculates an index for SI, but also quantitates other variables that are considered important in maintaining glucose tolerance, such as insulin secretion and the effect of glucose on its own disappearance (SG). Therefore, the minimal model provides a more comprehensive view of the glucose regulating system than a simple glucose tolerance test.
Knowledge of the influence of hypothyroidism on glucose metabolism may have important implications in successful glycemic control of diabetic dogs and perhaps in understanding the pathogenesis of diabetes mellitus in some dogs. The purpose of the study reported here was to evaluate whether hypothyroidism causes insulin resistance and to determine the overall effect of hypothyroidism on glucose tolerance in dogs. Additional objectives were to characterize the secretion profile of hormones that are counter-regulatory to insulin (ie, GH, IGF-1, and cortisol) and determine whether insulin resistance was associated with obesity in hypothyroid dogs.
Materials and Methods
Animals—Sixteen anestrous mixed-breed bitches, aged 25 to 39 months and weighing 9 to 13 kg, were used in the study. All dogs were determined healthy through findings of routine physical examination, CBC, serum biochemical analysis, heartworm antigen test, and parasitological fecal examination. Dogs were housed individually in controlled kennel conditions with 12 hours of light and 12 hours of dark and were fed a standard diet.a
Hypothyroidism was induced in 8 randomly selected dogs by IV administration of 1 mCi of iodine 131 (131I)/kg 37 to 58 weeks prior to the start of the study. Random selection was accomplished by an individual other than the investigators, who drew dog identification numbers from a container, with every other dog being assigned to the hypothyroid group. The remaining 8 dogs were used as euthyroid control dogs. Hypothyroidism was confirmed by ensuring a serum T4 concentration < 10 nmol/L before and 4 hours after IV administration of 50 μg of human recombinant TSHb 9 and 38 to 45 weeks after iodine 131 administration.19 All experiments were performed in anestrous dogs in the morning, after food had been withheld for 12 hours. All experimental procedures were approved by the Virginia Tech Animal Care and Use Committee.
Insulin-modified FSIGT—The insulin-modified FSIGT was performed as described elsewhere,20 with slight modifications. On the day of the experiment, 18-gaugec and 20-gauged indwelling catheters were placed in a jugular and cephalic vein, respectively. Dogs were allowed to rest quietly for 30 minutes in a cage and 5 minutes on a table before experiments began. Two milliliters of blood was obtained through the jugular catheter, which was flushed with 1 mL of heparinized (5 U/mL) saline (0.9% NaCl) solution before and after each sample collection point. The total dose of heparin did not exceed 600 U/dog. Blood samples were obtained at 20, 10, 5, and 1 minutes before (−20, −10, −5, and −1 minutes) and 2, 3, 4, 6, 8, 10, 12, 14, 16, 19, 22, 23, 24, 25, 27, 30, 40, 50, 60, 70, 80, 100, 120, 140, 160, and 180 minutes after IV administration (0 minutes) of 0.3 mg of dextrose/kg as a 25% solution in hypotonic saline (0.45% NaCl) solutione over 1 minute through the cephalic vein catheter. Human recombinant regular insulinf (0.03 U/kg) was administered through the cephalic catheter over 10 seconds at 20 minutes. Collected blood samples were immediately placed in chilled tubes coated with 5 mg of sodium fluoride and 4 mg of potassium oxalated and stored on ice. Within 30 minutes after collection, samples were centrifuged at 2,500 × g at 4°C for 12 minutes. Plasma was separated and frozen in aliquots at −70°C for later measurement of glucose and insulin concentrations.
Minimal model analysis—Basal plasma glucose and insulin concentrations were calculated as the mean values of the initial 4 samples obtained before glucose injections were administered. Insulin sensitivity and SG were calculated by use of a computer program,g as described elsewhere.21 The SI, defined as the fractional glucose uptake rate per unit of plasma insulin, represents the ability of insulin to enhance glucose uptake from plasma and to inhibit glucose release into plasma. Glucose effectiveness reflects the effect of glucose per se to enhance its own cellular uptake from plasma and to normalize glucose concentration independently of an increase in plasma insulin to higher than the basal concentration.
Acute insulin response to glucose was calculated as the AUC of the plasma insulin concentrations from 0 to 19 minutes after glucose administration. This index represents the mean insulin concentration higher than the basal concentration during the first phase of insulin secretion in response to glucose administration IV. The DI was calculated as the product of AIRG and SI. It represents a measure of overall glucose tolerance and is an index of β-cell function that describes the relation between insulin secretion and SI.22
Insulin and glucose assays—Plasma glucose concentrations were determined in duplicate with an automated chemistry analyzer by use of the hexokinase method.h A competitive binding, double-antibody radioimmunoassay for human insulini that has been validated for use in dogs was used according to the manufacturer's instructions. All samples were assayed in duplicate. The insulin radioimmunoassay was validated for canine plasma collected in tubes containing sodium fluoride and potassium oxalate by determining analytic sensitivity, precision, and linearity of dilution. Analytic sensitivity was determined by calculating the point of 95% of total binding on the standard curve. Interassay precision was determined by calculating the CV of 3 pools of canine plasma samples containing low, medium, and high concentrations of insulin in 7 assays performed on different dates. Intra-assay precision was evaluated by determining the CV from 10 replicates of each of 3 pools of plasma samples measured in 1 assay. Linearity of dilution was evaluated by measuring the insulin concentration in a plasma sample with high endogenous insulin concentration and in a sample obtained after exogenous insulin administration. Each sample was diluted with the insulin-free standard solution supplied with the kit at 1:2, 1:4, and 1:8 dilutions.
GH stimulation and suppression tests—Growth hormone stimulation and suppression tests were performed 1 week apart in 6 hypothyroid and 6 control dogs 78 to 110 weeks after hypothyroidism was induced. Blood samples were obtained via jugular venipuncture and placed in untreated collection tubesd at −15, 0, 15, 30, 45, 60, and 90 minutes after IV administration of GHRHj (1 μg/kg) for stimulation or somatostatinj (10 μg/kg) for suppression of GH through a cephalic venous catheter.23,24 Blood samples were allowed to clot for 20 minutes prior to centrifugation at 1,500 × g for 15 minutes, and serum was harvested and then stored at −70°C. In addition to the 2 basal samples for each test, 2 blood samples were collected from each dog 15 minutes apart on 2 separate days for measurement of serum GH concentration to be used for the calculation of the mean baseline GH concentration (8 samples). Mean serum concentrations of IGF-1 were calculated from 2 baseline samples, each obtained during 1 of the 2 dynamic GH tests.
Analysis of GH and IGF-1—Serum GH concentration was determined by use of a commercially available radioimmunoassay for canine and porcine GH.k Purified recombinant porcine GH standards (1 to 100 ng/mL), antiserum produced in guinea pigs inoculated with porcine GH, iodine 125 (125I)–labeled porcine GH, and a precipitating reagent complex (goat anti–guinea pig IgG serum in 3% polyethylene glycol and 0.05% nonionic surfactant in 0.05M PBSS, 0.025M EDTA, and 0.08% sodium azide) were provided in the kit, which was used without modification. The amino acid sequences of porcine and canine GH are identical.25
Intra-assay precision was determined by calculating the CV from 5 measurements (in duplicate) of GH in 3 pools of canine serum samples (containing 20.4, 7.7, or 1.5 ng of GH/mL) measured in 1 assay. Interassay CV was calculated from measurements of GH in the pooled samples in duplicate, 6 times during 1 month. Linearity was demonstrated by preparing aliquots of pooled canine serum samples containing various concentrations of GH and measuring them after dilution. Three pools of canine serum samples containing 7.7, 20.4, and 45.5 ng of GH/mL were measured by use of serum volumes of 100 (full volume), 75, 50, and 25 μL. Zero standard solution was added as required to raise the assay volume to 100 μL. The percentage of expected concentration was calculated for each pool. In addition, recovery was determined by adding various concentrations (0, 2.0, 5.0, and 20 ng/mL) of canine GHl to 3 pools of canine serum samples. Serum growth hormone concentration was determined in pools that were measured in duplicate 3 times in 1 assay. Percentage recovery was calculated by use of the observed versus expected concentration data.
Serum IGF-1 concentrations were measured in 7 control and 8 hypothyroid dogs by use of a commercially available radioimmunoassay.m The manufacturer of the kit reports 0.05% cross-reactivity with IGF-2 and ≤ 0.012% cross-reactivity with insulin, proinsulin, and C-peptide. The design of this assay involves measurement of IGF-1 in prediluted samples with a standard curve of great sensitivity (0 to 1.3 nmol/L standards). Samples are diluted in an acidification buffer to achieve dissociation of IGF-1 from binding proteins. The factor of dilution is dependent on range of values expected in the biological specimen. For physiologic concentrations of IGF-1 in serum, the manufacturer recommends a sample-to-acidification buffer ratio of 1:100. The assay begins when an aliquot of the acidified sample is mixed with a reagent containing first (anti–IGF-1) antibody and a physiologically excessive amount of IGF-2. This reagent neutralizes the acidic pH and the IGF-2 occupies IGF-binding proteins. Radiolabeled IGF-1 is then added to assay tubes to compete with endogenous IGF-1 for binding with assay antibody. After incubation, antibody-bound radioligand is isolated with second antibody, precipitating reagent, and centrifugation. The volumes of reagents and incubation times used were in accordance with the manufacturer's protocol.
Estimates of assay sensitivity, intra-assay precision, and interassay precision were obtained in a manner similar to that for insulin and GH assays. Linearity was assessed by comparing results from 3 serum samples diluted at various rates with acidification and dilution buffer. Recovery of IGF-1 was estimated by taking canine serum samples with various concentrations of IGF-1, mixing aliquots of these samples in various proportions (eg, 1:1, 1:2, and 1:3), and determining the concentration measured versus the concentration expected.
Urine cortisol concentrations—Two morning urine samples were collected via cystocentesis from 7 euthyroid and 8 hypothyroid dogs for urine cortisol-to-creatinine concentration ratio determinations. Each sample was collected by cystocentesis on separate days, at least 7 days apart from any other experiment. Urine for cortisol and creatinine measurements was frozen at −70°C for future assays by means of a previously validated method.26 Cortisol concentration was measured by use of a commercially available radioimmunoassay.n
DEXA scans—Eight hypothyroid dogs and 8 euthyroid dogs underwent a DEXA scan to determine total body mass, lean body mass, and absolute and percentage body fat as well as bone mineral density Dogs were anesthetized with isofluraneo only as part of an unrelated experiment conducted on the same day The DEXA scans were performed with dogs positioned in sternal recumbency (standard scan mode of 1.8 Gy).p In addition, a region of interest was drawn around the abdomen and the percentage of abdominal fat was calculated with the aid of software provided by the manufacturer.
Statistical analysis—Experimental data are reported as mean ± SD unless stated otherwise. Data were analyzed for normality and equal variances. Statistical comparison of means between the hypothyroid and control group was performed by use of the pooled-variance t test for equal variances or Welch-Satterthwaite t test for unequal variances. When necessary, data were logarithmically transformed to achieve normality. Correlation and simple regression analyses were performed for the data of each group to identify relationships between basal insulin, SI, and variables derived from the minimal model analysis and parameters from the GH analyses, urine cortisol-to-creatinine concentration ratios, serum IGF-1 concentrations, and data from the DEXA scan procedure. The Pearson correlation coefficient (r) was calculated for normally distributed data, whereas the Spearman rank correlation coefficient (ρ) was calculated for nonnormally distributed data.
Data from all 8 available basal samples/dog were used to calculate mean basal serum GH concentrations. Area under the curve was calculated for results of GH stimulation and suppression tests by applying the trapezoidal rule. Peak and nadir GH concentrations as well as differences from peak and nadir to baseline (absolute number and percentage) were also calculated. The means of each of those parameters were used to test for differences between groups by use of the appropriate t tests.
Statistical analyses were performed with a proprietary statistical program.q Values of P < 0.05 were considered significant for all analyses
Results
Dogs—All hypothyroid dogs had clinical signs of hypothyroidism, including weight gain, thin coat or alopecia, and lethargy. Although mean ± SD body weight was similar between hypothyroid (9.7 ± 1.18 kg) and control (euthyroid) groups (9.8 ± 0.77 kg) prior to induction of hypothyroidism, hypothyroid dogs (11.9 ± 0.59 kg) weighed significantly (P = 0.006) more than control dogs (10.4 ± 1.23 kg) at the time of the FSIGT. Serum T4 concentrations before and after TSH administration were < 5 nmol/L (lower limit of sensitivity of the T4 assay) in all hypothyroid dogs. In control dogs, all post–TSH administration serum T4 concentrations were > 35 nmol/L. The mean serum T4 concentration before and after TSH administration in control dogs was 26 ± 7 nmol/L and 57 ± 18 nmol/L, respectively.
Insulin assay validation—Sensitivity of the insulin assay was 1 μU/mL. The interassay CV for low (4 μU/mL), medium (19.1 μU/mL), and high (40 μU/mL) serum insulin concentrations was 12.3%, 11.1%, and 9%, respectively. The intra-assay CV for low (4 μU/mL), medium (20.6 μU/mL), and high (46.3 μU/mL) insulin concentrations was 18.4%, 6%, and 5.9%, respectively.
Percentage recovery of insulin after serial dilutions (1:2, 1:4, and 1:8) was 93%, 106%, and 118%, respectively, for the sample containing endogenous insulin only at an initial concentration of 59 μU/mL and 92%, 82%, and 90%, respectively, for the sample containing both canine and human recombinant insulin at an initial concentration of 119 μU/mL.
GH assay validation—Assay sensitivity was 1 ng/mL. The intra-assay CV calculated from repeated measurement of GH concentration in serum containing a high (20.4 ng/mL), medium (7.7 ng/mL), and low (1.5 ng/mL) GH concentration was 4.5%, 2.0%, and 10.4%, respectively. The interassay CV calculated from repeated measurement of GH concentration in the same serum pools containing high, medium, and low GH concentrations measured in 6 assays was 4.1%, 5.0%, and 9.1%, respectively.
To demonstrate completeness of recovery, various concentrations of canine GH were added to 3 pools of canine serum samples and the GH content was determined. Recovery pools were run 4 times in duplicate in 1 assay. The mean percentage of GH recovery was 94.4%, 98.9%, and 95.1%, respectively, for serum pools containing high, medium, and low GH concentration. Intra-assay CVs when canine GH (25 and 50 ng/mL) was added to canine serum were 4.8% and 4.9%, respectively. Linearity was determined by use of 1.33-, 2- and 4-fold dilution on 3 serum pools with various GH concentrations, and percentages of expected GH concentration were 92%, 94%, and 70%, respectively, for each dilution of the serum pool containing 7.7 ng/mL at full volume (100 μL); 96%, 99%, and 96%, respectively, for the sample containing 20.4 ng/mL; and 94%, 97%, and 99%, respectively, for the sample with 45.5 ng/mL of GH.
IGF-1 assay validation—For samples diluted at a ratio of 1:100, the analytic sensitivity of the assay, defined as the concentration of IGF-1 at 95% of total specific binding of IGF-1, was 8 nmol/L. Intra-assay CVs from 7 to 10 replicates of 3 canine serum samples representing the low (8 nmol/L [mixed-breed dog]), mid (36 nmol/L [Rhodesian Ridgeback]), and upper range (101 nmol/L [Mastiff]) were 10%, 7%, and 4%, respectively. Interassay CVs from the same low- and upper-range samples with a different midrange sample (50 nmol/L [Golden Retriever]) in 6 assays were 20%, 14%, and 12%, respectively. When the low-range sample was diluted with acidification buffer at ratios of 1:25, 1:50, and 1:100, the CV of the results was 11%. When the midrange (50 nmol/L) and the high-range (101 nmol/L) samples were assayed at dilutions of 1:50, 1:100, and 1:200, the CVs of results adjusted for dilution were both 9%. Canine serum samples with individual concentrations of 15 and 80 nmol/L were mixed at relative amounts of 1:1, 1:3, and 3:1, and 2 other samples with individual concentrations of 19 and 41 nmol/L were mixed at ratios of 1:1, 1:1, and 2:1 prior to assaying at a 1:100 dilution. The assay of the mixed samples consistently yielded 83% of the expected concentration of IGF-1.
Insulin-modified FSIGT and minimal model analysis—One dog in the control group was excluded from analysis because of prolonged moderate, subclinical hypoglycemia (< 50 mg/dL) after exogenous insulin administration. Seven dogs in the control group and 2 dogs in the hypothyroid group had a plasma glucose concentration < 70 mg/dL (58 ± 8.9 mg/dL). This was noticed once in 2 dogs and at 2 measurement points in 3 dogs, 3 measurement points in 3 dogs, and 4 measurement points in 1 dog.
Mean basal plasma insulin concentrations (Figure 1) were higher (8.2 ± 3.2 μU/mL and 4.2 ± 2.3 μU/mL, respectively; P = 0.019), whereas basal plasma glucose concentrations (Figure 2) were lower (90.1 ± 4.4 mg/dL and 96.4 ± 3.2 mg/dL, respectively; P = 0.008) in the hypothyroid group than in the control group. The AIRG was higher (P = 0.010) in the hypothyroid dogs (459 ± 258 μU/mL) than in the control dogs (159 μU/mL ± 56 μU/mL). Mean SI for hypothyroid dogs (4.9 ± 1.2 × 10−4 min−1/μU/mL) was lower (P < 0.001) than that for the control group (23.6 ± 5.3 × 10−4 min−1/μU/mL). In contrast, SG was similar (P = 0.977) between control (0.043 ± 0.009 min−1) and hypothyroid (0.043 ± 0.011 min−1) groups. The DI was lower in the hypothyroid group (2,244 ± 1,386 min−1) than in the control group (3,376 ± 1,616 min−1), but the difference was not significant (P = 0.070).
GH and IGF-1—No difference was evident between groups in regard to peak and nadir serum GH concentrations, differences from peak and nadir to baseline, and the AUC measurements in response to GHRH or somatostatin administration (Table 1), but individual dogs had marked variability in their responses to each hormone administered (Figures 3 and 4). On the other hand, basal serum GH and IGF-1 concentrations were significantly (P = 0.022 for GH and P = 0.029 for IGF-1) higher in the hypothyroid group than in the control group (Figure 5).
Mean ± SD basal serum GH and IGF-1 concentrations and AUCs for GHRH stimulation and somatostatin suppression tests in 8 anestrous mixed-breed bitches with experimentally induced hypothyroidism and 8 euthyroid control dogs.
Variable | Control dogs | Hypothyroid dogs | P value |
---|---|---|---|
Basal GH (ng/mL) | 2.24 ± 1.35 | 3.98 ± 1.25 | 0.022 |
GNRH stimulation test | |||
AUC (min·ng/mL) | 159.25 ± 107.42 | 264.81 ± 205.87 | 0.29 |
Peak GH (ng/mL) | 3.05 ± 2.52 | 5.80 ± 5.78 | 0.32 |
Difference from peak to baseline (ng/mL) | 1.85 ± 2.22 | 2.38 ± 4.58 | 0.80 |
Somatostatin suppression test | |||
AUC (min·ng/mL) | 116.56 ± 75.04 | 214.94 ± 110.92 | 0.10 |
Nadir for GH (ng/mL) | 0.98 ± 0.86 | 2.02 ± 1.31 | 0.14 |
Difference from nadir to baseline (ng/dL) | −0.48 ± 0.41 | −1.16 ± 0.80 | 0.09 |
Basal IGF-1 (nmol/L) | 11.14 ± 4.63 | 17.56 ± 5.39 | 0.029 |
Urine cortisol-to-creatinine concentration ratio | 0.012 ± 0.007 | 0.014 ± 0.075 | 0.59 |
Urine cortisol-to-creatinine concentration ratios—No difference in cortisol-to-creatinine concentration ratio was found between the hypothyroid (0.014 ± 0.075) and control (0.012 ± 0.007) groups (P = 0.59).
DEXA—Mean total body weight was higher (P = 0.006) in the hypothyroid group (11.9 ± 0.6 kg), compared with euthyroid dogs (10.4 ± 1.2 kg), but there were no differences in lean body mass (6.3 ± 0.7 kg vs 6.6 ± 1.0 kg, respectively), total absolute body fat (3.7 ± 1.5 kg vs 4.9 ± 1.1 kg, respectively), and percentage body fat (36.1 ± 11.1% vs 42.4 ± 9.0%, respectively) of control versus hypothyroid dogs. Hypothyroid dogs had a relative increase (P = 0.007) in abdominal body fat (47.5 ± 3.6%), compared with control dogs (34.6 ± 8.8%).
Correlation analysis—No significant correlation was found between any of the variables tested (data not shown).
Discussion
The results of the present study suggested that hypothyroidism causes substantial insulin resistance, as evidenced by an almost 5-fold decrease in SI, compared with the SI in euthyroid dogs. Moreover, results derived from minimal model analysis showed that hypothyroid dogs are able to maintain glucose tolerance by a compensatory increase in insulin secretion and by maintaining clinically normal SG. These findings expand on those of previous studies in dogs with glucose intolerance but only inferred insulin resistance. Through use of standard glucose tolerance tests, hypothyroid dogs have been found to have exaggerated insulin responses, clinically normal or higher than usual plasma glucose concentrations, and slower than typical glucose disappearance,4–6 which is consistent with our observations.
To better understand the mechanisms leading to altered carbohydrate metabolism, we used the FSIGT with minimal model analysis, which was developed to allow a dynamic approach to the assessment of SI and other factors that affect glucose metabolism.27,28 The minimal model of glucose and insulin kinetics represents the least complex mathematical model that relates to the effect of glucose itself (measured as SG) and the effect of insulin (SI) to promote glucose uptake into cells and inhibit its production and release by the liver (HGO). A computer program for the nonlinear least squares estimation method searches for the best fit of insulin and glucose data, yielding parameters of SI, SG, and first-phase pancreatic responsivity (AIRG).21 Findings reported here are in agreement with studies in humans9,10,29–31 and rodents8,32 that hypothyroidism causes insulin resistance.
In addition to the decrease in SI, hypothyroid dogs in the present study had no change in SG, increased AIRG relative to that in euthyroid dogs, and no change in DI. Glucose effectiveness, expressed as the fractional disappearance rate of glucose at basal plasma insulin concentrations, can be explained as the ability of glucose per se to enhance its own uptake and, to a lesser degree, to suppress endogenous glucose production by the liver, independent of insulin.33–35 The effect of SG is at least as important as the dynamic insulin response to maintain clinically normal blood glucose concentrations and thus glucose tolerance in dogs.33,34 In humans with non–insulin dependent diabetes mellitus, SG is unaffected despite profound insulin resistance.36,37 Because insulin resistance in non–insulin dependent diabetes mellitus is associated with inadequate insulin secretion, SG has a primary role in overall glucose disposal. In the present study, values obtained for SG were virtually identical between groups, indicating that hypothyroidism had no effect on this variable. However, because insulin resistance reduces glucose disposal at insulin-sensitive sites, SG may be of greater importance in hypothyroid versus euthyroid dogs, particularly when concurrent diabetes mellitus is present.
The AIRG in the study dogs was nearly 3-fold as high in the hypothyroid group, compared with the AIRG in the euthyroid group, compensating for the insulin resistance. The magnitude of change in secretion accounts for the lack of change in the DI in hypothyroidism, which is derived from insulin secretion and sensitivity (AIRG × SI). This indicated that the compensatory insulin secretory response was sufficient to sustain normal glucose tolerance. The DI reflects a fundamental concept in the understanding of beta-cell function and indicates that any change in one variable of glucose tolerance in a healthy individual is mirrored by a reciprocal change in the other variable, otherwise hyperglycemia develops.38 When this concept is applied to the results of the present study, the suggestion is that hypothyroid dogs have sufficient beta-cell function to remain glucose tolerant and euglycemic. Our results were consistent with clinical observations in dogs with concurrent spontaneous hypothyroidism and diabetes mellitus, in which insulin secretion was insufficient to compensate for the insulin resistance, resulting in poor glycemic regulation.1 As glycemic control improved with treatment of hypothyroidism in those dogs, insulin resistance improved, in a manner similar to that in hypothyroid humans.9,10
Despite the fact that chronic stimulation of insulin secretion can lead to pancreatic beta-cell exhaustion and that hypothyroidism has been identified as a risk factor for pancreatitis in dogs, the evidence linking hypothyroidism and diabetes mellitus is limited to the finding of diabetes mellitus in 10% of hypothyroid dogs in 1 retrospective study2 and hypothyroidism as a concurrent disease in 4% of dogs in a retrospective study3 of diabetes mellitus. Additional studies are necessary to determine whether hypothyroidism contributes to development of diabetes mellitus in dogs.
The cause of the markedly impaired SI in the hypothyroid dogs of the present study is likely multifactorial. Lower than typical glucose uptake in insulin-sensitive tissues including muscle and fat could be explained by a decreased expression of the insulin-sensitive GLUT4. Several studies have shown a stimulatory effect of thyroid hormones on GLUT4 expression leading to increased glucose and uptake39–42 and the presence of a triiodothyronine-responsive region within the GLUT4 promoter,43 whereas lower than typical GLUT4 expression has been detected in monocytes of hypothyroid humans with insulin resistance.30 Low numbers of insulin receptors and low binding affinity have been detected in adipose tissue of hypothyroid humans as well as in adipose, muscle, and liver tissues of hypothyroid rats in some studies29,31,44,45 and may contribute to decreased SI.
Postreceptor alterations in response to insulin, resulting in a decrease in glycolysis and glucose oxidation in skeletal muscle, may play an additional role in the development of insulin resistance.8,29,31,32,46 Hypothyroidism reduces glucose use secondary to marked impairment of the vasodilatory response to insulin that occurs in the postprandial state in skeletal muscle and adipose tissue.47 In fact, the decrease in insulin-stimulated glucose uptake in hypothyroid humans is largely accounted for by the decrease in insulin delivery due to reduced forearm blood flow, indicating that reduced endothelium-dependent vasodilation plays an important role in insulin action in hypothyroidism.
A decrease in HGO could account for the finding of lower, albeit clinically normal, basal plasma glucose concentrations in the hypothyroid dogs of the present study. Hepatic gluconeogenesis48–50 and rates of glycogenolysis and glucose release in response to insulin51 are diminished in hypothyroid rats and rabbits, respectively. Results of more recent studies9,52 also suggest that hypothyroid humans have a decrease in HGO. Together, these findings indicate that hypothyroidism does not negatively affect hepatic SI and HGO and that insulin resistance at peripheral sites may be, at least in part, counterbalanced by a decrease in glucose output by the liver, thereby maintaining euglycemia. Future FSIGT studies including tritium (3H)-labeled glucose would allow identification of the contribution of cellular glucose utilization and endogenous glucose production on insulin resistance53 in hypothyroidism.
The finding of a lower than typical basal plasma glucose concentration in hypothyroid dogs illustrates that mechanisms other than hyperglycemia must be the stimulus for the increased basal insulin concentrations. A decrease in insulin clearance could account for an increase in peripheral insulin concentration, as has been reported for insulin-resistant, hypothyroid women10 and insulin-resistant obese dogs, in which beta-cell hypersecretion co-occurs.54 In addition, higher than typical insulin clearance and lower than typical insulin half-life have been reported for hypothyroid humans.55,56 Additional studies of hypothyroid dogs are needed to assess the differential roles of insulin secretion and clearance and hepatic extraction on plasma insulin concentrations so that their relative contributions to the increased plasma insulin concentrations found in the present study can be determined. Furthermore, evaluation of skeletal muscle and adipose tissue biopsy specimens from hypothyroid dogs for GLUT4 could help elucidate the pathogenesis of insulin resistance.
In the present study, the insulin-modified FSIGT was chosen instead of the original FSIGT because of the advantage it offers for more accurate modeling in situations in which insulin secretion is insufficient. Mild hypoglycemia is common when this test is used in humans and dogs,57,58 as we noted in some study dogs. Recently, an artifactual decrease in SI was identified in clinically normal humans by use of the insulin-modified FSIGT, which induced hypoglycemia, compared with results of the same test protocol when hypoglycemia was prevented by infusion of glucose.59 The decision was made to exclude 1 dog from the euthyroid group in the present study because of the marked and prolonged hypoglycemia. Given that hypoglycemia was more prevalent in the control group, any reduction hypoglycemia induced in SI would have lessened the difference between the hypothyroid and euthyroid groups, making the finding of insulin resistance in hypothyroid dogs even more profound.
In addition to the direct effects of hypothyroidism on SI, the disease may alter other factors important in glucose homeostasis. As in dogs with naturally occurring hypothyroidism, of which approximately 40% are obese,2,60 dogs in the present study gained 23% of their body weight after induction of hypothyroidism. Obesity is a cause of insulin resistance in many species, including dogs.17,61 The DEXA evaluation of the abdominal fat region indicated the hypothyroid dogs in the study had substantial visceral adiposity or central obesity. This type of obesity has a strong association with insulin resistance in euthyroid humans.61,62 The reason for the association is not entirely clear, but an increase in plasma free fatty acids and alterations in adipocytokines and inflammatory cytokines have been proposed as possible mediators. Although these factors were not evaluated in the present study, free fatty acids are reportedly unchanged or decreased in hypothyroid dogs5,7; therefore, their role in hypothyroid-mediated insulin resistance is unclear.
Serum leptin concentration, which is the only adipocytokine evaluated in hypothyroid dogs to the authors' knowledge, can increase concurrently with a high serum insulin concentration, even after adjusting for body condition score in hypothyroid dogs.63 However, the clinical importance of this is unclear, particularly because body condition scoring does not accurately detect central obesity. The insulin resistance in hypothyroid dogs reported here was marked, compared with that previously observed in dogs with experimentally induced obesity,17,54 so it is likely that the contribution of abdominal adiposity to insulin resistance does not solely account for the magnitude of insulin resistance in hypothyroidism.
Because an excess in circulating GH can cause considerable impairment of glucose homeostasis in dogs, alterations in control and serum concentrations of this hormone and of IGF-1 as a marker of its tissue activity were evaluated in the present study. The high basal serum GH and IGF-1 concentrations in hypothyroid dogs were consistent with findings of other studies14–16 in this species. Serum GH concentration is reportedly high in hypothyroid humans with insulin resistance,10 but other studies64–66 in hypothyroid humans have revealed no change or a decrease in serum GH concentrations or response to GHRH. Hypothyroid humans also have consistently been shown to have a decrease in serum IGF-1 concentration.67–69 Similar decreases in GH and IGF-1 concentration have been reported for hypothyroid rodents as well.70,71
Dogs appear to be unique in having an increase in circulating GH and IGF-1, compared with other species that have a decrease or no change in these hormones with hypothyroidism. It has been suggested that pituitary thyrotrophs undergo transdifferentiation to secrete GH as well as thyrotropin.14 However, because GH and IGF-1 concentrations in the present study were not correlated with indices of SI, their contribution to insulin resistance in hypothyroid dogs is unknown.
The reason the responses to both GHRH and somatostatin administration were so variable (Figures 3 and 4) is also unclear, but this variability prevented significant differences from being identified despite means of most of the measurements suggesting an excess in circulating GH in hypothyroid dogs (Table 1). Circulating concentrations of other counter-regulatory hormones are reportedly altered by hypothyroidism in humans, including cortisol.10 The effects of hypothyroidism on plasma and urine concentrations of cortisol and cortisol metabolites in humans are inconsistent,10,72–74 and no difference in urine cortisol-to-creatinine concentration ratios was evident between hypothyroid and control dogs in the present study.
Hypothyroid dogs appear to be markedly insulin resistant, but glucose tolerance is unaffected as a consequence of an increase in insulin secretion and unchanged SG. Although hypothyroidism alone does not cause overt clinical disturbances in glucose homeostasis, knowledge of the status of thyroid gland function could have important implications in the successful management of diabetes mellitus in dogs. In dogs with diabetes mellitus that is challenging to manage and in which insulin resistance is suspected, hypothyroidism should be included in the list of differential diagnoses, with correction thereof potentially leading to better glycemic control. The pathogenesis of insulin resistance appears to be multifactorial, with high serum GH and IGF-1 concentrations potentially contributing. In addition, the role of visceral fat on glucose metabolism may also be important in the development of insulin resistance in hypothyroid dogs.
ABBREVIATIONS
AIRG | Acute insulin response to glucose |
AUC | Area under the curve |
CV | Coefficient of variation |
DEXA | Dual energy x-ray absorptiometry |
DI | Disposition index |
FSIGT | Frequently sampled IV glucose tolerance test |
GH | Growth hormone |
GHRH | Growth hormone–releasing hormone |
GLUT | Glucose transporter |
HGO | Hepatic glucose output |
IGF | Insulin-like growth factor |
SG | Glucose effectiveness |
SI | Insulin sensitivity |
T4 | Thyroxine |
TSH | Thyroid-stimulating hormone |
Hill's Science Diet adult dry kibble, Hill's Pet Nutrition Inc, Topeka, Kan.
Thyrogen, Genzyme Corp, Framingham, Mass.
Venocath, Abbott Laboratories, Abbott Park, Ill.
Becton Dickinson, Franklin Lakes, NJ.
Vedco Inc, St Joseph, Mo.
Humulin R, Eli Lilly, Lake Forest, Ill.
MINMOD Millenium, version 6.02, Richard Bergman, University of Southern California, Los Angeles, Calif.
Olympus AU 400 Chemistry Analyzer and OSR6121 Glucose Reagent, Olympus America Inc, Melville, NY.
DSL-1600 insulin radioimmunoassay, Diagnostic Systems Laboratories Inc, Webster, Tex.
Bachem Bioscience Inc, King of Prussia, Pa.
PGH-46HK, Linco Research, St Charles, Mo.
Provided by Dr. AF Parlow, Director, Pituitary Hormones and Antisera Center, Harbour-UCLA Medical Center, Torrance, Calif.
IGF-1 RIA, Mediagnost, Reutlingen, Germany.
Coat-a-count Cortisol, Siemens Medical Solutions Diagnostics, Los Angeles, Calif.
Abbott Laboratories, Abbott Park, Ill.
Lunar Prodigy Advance, GE Healthcare, Madison, Wis.
SAS, version 9.1.3, SAS Institute Inc, Cary, NC.
References
- 1.↑
Ford SL, Nelson RW, Feldman EC, et al. Insulin resistance in three dogs with hypothyroidism and diabetes mellitus. J Am Vet Med Assoc 1993; 202:1478–1480.
- 2.↑
Dixon RM, Reid SW, Mooney CT. Epidemiological, clinical, haematological and biochemical characteristics of canine hypothyroidism. Vet Rec 1999; 145:481–487.
- 3.↑
Hess RS, Saunders HM, Van Winkle TJ, et al. Concurrent disorders in dogs with diabetes mellitus: 221 cases (1993–1998). J Am Vet Med Assoc 2000; 217:1166–1173.
- 4.
Renauld A, Pinto JE, Sverdlik RC, et al. Studies on the effect of thyroidectomy on the insulin response to hyperglycaemia in the dog. Acta Physiol Lat Am 1973; 23:29–36.
- 5.
Renauld A, Sverdlik RC, Andrade LL. Effects of hypothyroidism on serum immunoreactive insulin, free fatty acids and blood sugar in the dog as tested for oral glucose tolerance. Corrective effects of thyroxine therapy. Horm Metab Res 1974; 6:137–141.
- 6.
Renauld A, Sverdlik RC, Andrade LL, et al. Studies on the effect of thyroxine replacement therapy on the increased insulin response to hyperglycemia in the thyroidectomized dog. Horm Metab Res 1972; 4:373–376.
- 7.
Renauld A, Sverdlik RC & Garrido D. Blood sugar, serum insulin and serum free fatty acid responses to graded glucose pulses in hypothyroid dogs. Acta Diabetol Lat 1982; 19:97–105.
- 8.
Cettour-Rose P, Theander-Carrillo C, Asensio C, et al. Hypothyroidism in rats decreases peripheral glucose utilisation, a defect partially corrected by central leptin infusion. Diabetologia 2005; 48:624–633.
- 9.
Rochon C, Tauveron I, Dejax C, et al. Response of glucose disposal to hyperinsulinaemia in human hypothyroidism and hyperthyroidism. Clin Sci (Lond) 2003; 104:7–15.
- 10.↑
Stanicka S, Vondra K, Pelikanova T, et al. Insulin sensitivity and counter-regulatory hormones in hypothyroidism and during thyroid hormone replacement therapy. Clin Chem Lab Med 2005; 43:715–720.
- 11.
Eigenmann JE, Eigenmann RY, Rijnberk A, et al. Progesterone-controlled growth hormone overproduction and naturally occurring canine diabetes and acromegaly. Acta Endocrinol (Copenh) 1983; 104:167–176.
- 12.
Hurty CA & Flatland B. Feline acromegaly: a review of the syndrome. J Am Anim Hosp Assoc 2005; 41:292–297.
- 13.
Møller N, Jørgensen JO. Effects of growth hormone on glucose, lipid, and protein metabolism in human subjects. Endocr Rev 2009; 30:152–177.
- 14.↑
Diaz-Espineira MM, Galac S, Mol JA, et al. Thyrotropin-releasing hormone-induced growth hormone secretion in dogs with primary hypothyroidism. Domest Anim Endocrinol 2008; 34:176–181.
- 15.
Diaz-Espineira MM, Mol JA, Rijnberk A, et al. Adenohypophyseal function in dogs with primary hypothyroidism and nonthyroidal illness. J Vet Intern Med 2009; 23:100–107.
- 16.
Lee WM, Diaz-Espineira M, Mol JA, et al. Primary hypothyroidism in dogs is associated with elevated GH release. J Endocrinol 2001; 168:59–66.
- 17.↑
Ellmerer M, Hamilton-Wessler M, Kim SP, et al. Reduced access to insulin-sensitive tissues in dogs with obesity secondary to increased fat intake. Diabetes 2006; 55:1769–1775.
- 18.↑
Bergman RN, Prager R, Volund A, et al. Equivalence of the insulin sensitivity index in man derived by the minimal model method and the euglycemic glucose clamp. J Clin Invest 1987; 79:790–800.
- 19.↑
Sauve F & Paradis M. Use of recombinant human thyroid-stimulating hormone for thyrotropin stimulation test in euthyroid dogs. Can Vet J 2000; 41:215–219.
- 20.↑
Tobin BL, Finegood DT. Reduced insulin secretion by repeated low doses of STZ impairs glucose effectiveness but does not induce insulin resistance in dogs. Diabetes 1993; 42:474–483.
- 21.↑
Pacini G, Bergman RN. MINMOD: a computer program to calculate insulin sensitivity and pancreatic responsivity from the frequently sampled intravenous glucose tolerance test. Comput Methods Programs Biomed 1986; 23:113–122.
- 22.↑
Mari A, Ahren B & Pacini G. Assessment of insulin secretion in relation to insulin resistance. Curr Opin Clin Nutr Metab Care 2005; 8:529–533.
- 23.
Fracassi F, Gandini G, Diana A, et al. Acromegaly due to a somatroph adenoma in a dog. Domest Anim Endocrinol 2007; 32:43–54.
- 24.
Meij BP, Mol JA, Hazewinkel HA, et al. Assessment of a combined anterior pituitary function test in Beagle dogs: rapid sequential intravenous administration of four hypothalamic releasing hormones. Domest Anim Endocrinol 1996; 13:161–170.
- 25.↑
Secchi C, Berrini A, Gaggioli D, et al. Amino acid modifications in canine, equine and porcine pituitary growth hormones, identified by peptide-mass mapping. J Chromatogr B Biomed Sci Appl 2001; 757:237–245.
- 26.↑
Randolph JF, Toomey J, Center SA, et al. Use of the urine cortisol-to-creatinine ratio for monitoring dogs with pituitary-dependent hyperadrenocorticism during induction treatment with mitotane (o, p'-DDD). Am J Vet Res 1998; 59:258–261.
- 27.
Bergman RN, Ider YZ, Bowden CR, et al. Quantitative estimation of insulin sensitivity. Am J Physiol 1979; 236:E667–E677.
- 28.
Toffolo G, Bergman RN, Finegood DT, et al. Quantitative estimation of beta cell sensitivity to glucose in the intact organism: a minimal model of insulin kinetics in the dog. Diabetes 1980; 29:979–990.
- 29.
Arner P, Bolinder J, Wennlund A, et al. Influence of thyroid hormone level on insulin action in human adipose tissue. Diabetes 1984; 33:369–375.
- 30.↑
Maratou E, Hadjidakis DJ, Kollias A, et al. Studies of insulin resistance in patients with clinical and subclinical hypothyroidism. Eur J Endocrinol 2009; 160:785–790.
- 31.
Pedersen O, Richelsen B, Bak J, et al. Characterization of the insulin resistance of glucose utilization in adipocytes from patients with hyper- and hypothyroidism. Acta Endocrinol (Copenh) 1988; 119:228–234.
- 32.
Dimitriadis G, Parry-Billings M, Bevan S, et al. The effects of insulin on transport and metabolism of glucose in skeletal muscle from hyperthyroid and hypothyroid rats. Eur J Clin Invest 1997; 27:475–483.
- 33.
Ader M, Pacini G, Yang YJ, et al. Importance of glucose per se to intravenous glucose tolerance. Comparison of the minimal-model prediction with direct measurements. Diabetes 1985; 34:1092–1103.
- 34.
Ader M, Ni TC, Bergman RN. Glucose effectiveness assessed under dynamic and steady state conditions. Comparability of uptake versus production components. J Clin Invest 1997; 99:1187–1199.
- 35.
Best JD, Kahn SE, Ader M, et al. Role of glucose effectiveness in the determination of glucose tolerance. Diabetes Care 1996; 19:1018–1030.
- 36.
Alzaid AA, Dinneen SF, Turk DJ, et al. Assessment of insulin action and glucose effectiveness in diabetic and nondiabetic humans. J Clin Invest 1994; 94:2341–2348.
- 37.
Baron AD, Kolterman OG, Bell J, et al. Rates of noninsulin-mediated glucose uptake are elevated in type II diabetic subjects. J Clin Invest 1985; 76:1782–1788.
- 38.↑
Ahren B & Pacini G. Importance of quantifying insulin secretion in relation to insulin sensitivity to accurately assess beta cell function in clinical studies. Eur J Endocrinol 2004; 150:97–104.
- 39.
Romero R, Casanova B, Pulido N, et al. Stimulation of glucose transport by thyroid hormone in 3T3-L1 adipocytes: increased abundance of GLUT1 and GLUT4 glucose transporter proteins. J Endocrinol 2000; 164:187–195.
- 40.
Weinstein SP, O'Boyle E, Haber RS. Thyroid hormone increases basal and insulin-stimulated glucose transport in skeletal muscle. The role of GLUT4 glucose transporter expression. Diabetes 1994; 43:1185–1189.
- 41.
Weinstein SP, Watts J, Haber RS. Thyroid hormone increases muscle/fat glucose transporter gene expression in rat skeletal muscle. Endocrinology 1991; 129:455–464.
- 42.
Torrance CJ, deVente JE, Jones JP, et al. Effects of thyroid hormone on GLUT4 glucose transporter gene expression and NIDDM in rats. Endocrinology 1997; 138:1204–1214.
- 43.↑
Richardson JM, Pessin JE. Identification of a skeletal muscle-specific regulatory domain in the rat GLUT4/muscle-fat gene. J Biol Chem 1993; 268:21021–21027.
- 44.
Czech MP, Malbon CC, Kerman K, et al. Effect of thyroid status on insulin action in rat adipocytes and skeletal muscle. J Clin Invest 1980; 66:574–582.
- 45.
Mackowiak P, Ginalska E, Nowak-Strojec E, et al. The influence of hypo- and hyperthyreosis on insulin receptors and metabolism. Arch Physiol Biochem 1999; 107:273–279.
- 46.
Dimitriadis GD, Leighton B, Parry-Billings M, et al. Effects of hypothyroidism on the sensitivity of glycolysis and glycogen synthesis to insulin in the soleus muscle of the rat. Biochem J 1989; 257:369–373.
- 47.↑
Dimitriadis G, Mitrou P, Lambadiari V, et al. Insulin action in adipose tissue and muscle in hypothyroidism. J Clin Endocrinol Metab 2006; 91:4930–4937.
- 48.
Comte B, Vidal H, Laville M, et al. Influence of thyroid hormones on gluconeogenesis from glycerol in rat hepatocytes: a dose-response study. Metabolism 1990; 39:259–263.
- 49.
Huang MT, Lardy HA. Effects of thyroid states on the Cori cycle, glucose-alanine cycle, and futile cycling of glucose metabolism in rats. Arch Biochem Biophys 1981; 209:41–51.
- 50.
Okajima F & Ui M. Metabolism of glucose in hyper- and hypothyroid rats in vivo. Relation of catecholamine actions to thyroid activity in controlling glucose turnover. Biochem J 1979; 182:585–592.
- 52.
Handisurya A, Pacini G, Tura A, et al. Effects of T4 replacement therapy on glucose metabolism in subjects with subclinical (SH) and overt hypothyroidism (OH). Clin Endocrinol (Oxf) 2008; 69:963–969.
- 53.↑
Cobelli C, Pacini G, Toffolo G, et al. Estimation of insulin sensitivity and glucose clearance from minimal model: new insights from labeled IVGTT. Am J Physiol 1986; 250:E591–E598.
- 54.↑
Mittelman SD, Van Citters GW, Kim SP, et al. Longitudinal compensation for fat-induced insulin resistance includes reduced insulin clearance and enhanced beta-cell response. Diabetes 2000; 49:2116–2125.
- 55.
Dimitriadis G, Baker B, Marsh H, et al. Effect of thyroid hormone excess on action, secretion, and metabolism of insulin in humans. Am J Physiol 1985; 248:E593–E601.
- 56.
Randin JP, Tappy L, Scazziga B, et al. Insulin sensitivity and exogenous insulin clearance in Graves' disease. Measurement by the glucose clamp technique and continuous indirect calorimetry. Diabetes 1986; 35:178–181.
- 57.
Finegood DT & Tzur D. Reduced glucose effectiveness associated with reduced insulin release: an artifact of the minimal-model method. Am J Physiol 1996; 271:E485–E495.
- 58.
Pacini G, Tonolo G, Sambataro M, et al. Insulin sensitivity and glucose effectiveness: minimal model analysis of regular and insulin-modified FSIGT. Am J Physiol 1998; 274:E592–E599.
- 59.↑
Brehm A, Thomaseth K, Bernroider E, et al. The role of endocrine counterregulation for estimating insulin sensitivity from intravenous glucose tolerance tests. J Clin Endocrinol Metab 2006; 91:2272–2278.
- 60.
Panciera DL. Hypothyroidism in dogs: 66 cases (1987–1992). J Am Vet Med Assoc 1994; 204:761–767.
- 61.
Mittelman SD, Van Citters GW, Kirkman EL, et al. Extreme insulin resistance of the central adipose depot in vivo. Diabetes 2002; 51:755–761.
- 62.
Kong A P, Chan NN, Chan JC. The role of adipocytokines and neurohormonal dysregulation in metabolic syndrome. Curr Diabetes Rev 2006; 2:397–407.
- 63.↑
Mazaki-Tovi M, Feuermann Y, Segev G, et al. Increased serum leptin and insulin concentrations in canine hypothyroidism. Vet J 2008; 183:109–114.
- 64.
Chernausek SD & Turner R. Attenuation of spontaneous, nocturnal growth hormone secretion in children with hypothyroidism and its correlation with plasma insulin-like growth factor I concentrations. J Pediatr 1989; 114:968–972.
- 65.
Hanew K, Utsumi A, Sugawara A, et al. Enhanced plasma GH responses to simultaneous administration of TRH and GHRH in patients with primary hypothyroidism. Endocr J 1995; 42:43–47.
- 66.
Valcavi R, Dieguez C, Preece M, et al. Effect of thyroxine replacement therapy on plasma insulin-like growth factor 1 levels and growth hormone responses to growth hormone releasing factor in hypothyroid patients. Clin Endocrinol (Oxf) 1987; 27:85–90.
- 67.
Akin F, Yaylali GF, Turgut S, et al. Growth hormone/insulin-like growth factor axis in patients with subclinical thyroid dysfunction. Growth Horm IGF Res 2009; 19:252–255.
- 68.
Miell JP, Zini M, Quin JD, et al. Reversible effects of cessation and recommencement of thyroxine treatment on insulin-like growth factors (IGFs) and IGF-binding proteins in patients with total thyroidectomy. J Clin Endocrinol Metab 1994; 79:1507–1512.
- 69.
Schmid C, Zwimpfer C, Brandle M, et al. Effect of thyroxine replacement on serum IGF-I, IGFBP-3 and the acid-labile sub-unit in patients with hypothyroidism and hypopituitarism. Clin Endocrinol (Oxf) 2006; 65:706–711.
- 70.
Burstein PJ, Draznin B, Johnson CJ, et al. The effect of hypothyroidism on growth, serum growth hormone, the growth hormone-dependent somatomedin, insulin-like growth factor, and its carrier protein in rats. Endocrinology 1979; 104:1107–1111.
- 71.
Ramos S, Goya L, Alvarez C, et al. Effect of thyroxine administration on the IGF/IGF binding protein system in neonatal and adult thyroidectomized rats. J Endocrinol 2001; 169:111–122.
- 72.
Hoshiro M, Ohno Y, Masaki H, et al. Comprehensive study of urinary cortisol metabolites in hyperthyroid and hypothyroid patients. Clin Endocrinol (Oxf) 2006; 64:37–45.
- 73.
Iranmanesh A, Lizarralde G, Johnson ML, et al. Dynamics of 24-hour endogenous cortisol secretion and clearance in primary hypothyroidism assessed before and after partial thyroid hormone replacement. J Clin Endocrinol Metab 1990; 70:155–161.
- 74.
Vantyghem MC, Ghulam A, Hober C, et al. Urinary cortisol metabolites in the assessment of peripheral thyroid hormone action: overt and subclinical hypothyroidism. J Endocrinol Invest 1998; 21:219–225.